Oberauer & Kliegl (2006) used nlme() of the nlme package to fit accuracy in various working-memory tasks as a function of presentation time and experimental conditions, most importantly memory load. In 2010, I checked for a small subset of our data whether such models can also be implemented with nlmer() of the lme4 package. The "problem" has been that the nlmer() interface is not quite as convenient and flexible as the nlme() interface for the specification of within-subject fixed effects and associated variance components and correlation parameters. Thus, these effects and their associated gradient parameters must be hand-coded for different models. Therefore, we have stayed with nlme() for our current research; see, e.g., Göthe et al. (2012) under Research Topics/Working Memory (TAFs, and NLMMs) for data and R scripts at this site.
Update 01-01-2013: Ben Bolker ( http://rpubs.com/bbolker/3363 ) shows how such functions can be set up more elegantly with the help of derive(), deparse(), and eval(), but it still is a bit of a hack anyway. My last checks of the nlmer-models were based on lme4_0.999999-0.
 Oberauer.Kliegl.JML.55.2006.601.pdf | [ ] | [01] | 812 Kb | 01/01/2013 12:38 |
 p1.nlme.R | [ ] | [02] | 2 Kb | 01/01/2013 12:31 |
| p1.nlmer.R | [ ] | [03] | 2 Kb | 01/01/2013 12:32 |
| Ftaf.R | [ ] | [04] | 1 Kb | 01/01/2013 12:32 |
| Ftaf.V2.R | [ ] | [05] | 0.5 Kb | 01/01/2013 12:33 |
| Ftaf.V3.R | [ ] | [06] | 0.6 Kb | 01/01/2013 12:34 |
| Ftaf.V4.R | [ ] | [07] | 0.6 Kb | 01/01/2013 12:34 |
| exampleTAF.rda | [ ] | [08] | 1 Kb | 01/01/2013 12:35 |
